ML20153D602
| ML20153D602 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 08/25/1988 |
| From: | Gridley R TENNESSEE VALLEY AUTHORITY |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| NUDOCS 8809020257 | |
| Download: ML20153D602 (44) | |
Text
'
't' g, F
( ' **
TENNESSEE VALLEY AUTHORITY CH ATTANOOGA. TENNESSEE 37401 SN 157B Lookout Place AUG 251988 U.S. Nuclear Regulatory Consnission ATTN: Document Control Desk Washington, D.C.
20552 Centlemen:
In the Matter of
)
Docket Nos.
50-259 Tennessee Valley Authority
)
50-260 50-296 BROWNS FERRY NUCLEAR PLANT (BFN) - PROBABILISTIC RISK ASSESSMENT (PRA) -
SUMMARY
REPORT
References:
1.
C. E. Cears' summary of May 16, 1988 meeting with the
ennessee Valley Authority dated May 27, 1988 2.
S. D. Ebneter's letter to S. A. White transmitting NRC staff cotxnents on the January 1986 version of the BFN PRA, dated October 1, 1987 This letter transmits the information requested by reference 1, fulfilling the commitment to provide a summary document by August 30, 1988. As requested, the report includes:
'The rationale for concluding that the revised PRA will conservatively reflect the configuration of Unit 2 at the time of re:-tart and
- A summary of the changes made between tho January 1986 PRA reviewed by the staff in its October 1, 1987 letter (reference 2) and the September 1987 version.
T6 purpose of the summary report is to provide addLLional information to support a concluslon that BFN is not an outiler with respect to severe accident characteristics when compared to plants of similar type and vintage.
No new commitments are made with this transmittal.
If you have any questions, please call D. L. Williams at (615) 632-7170.
Very truly yours, TENNESSEE VALLEY AUTHORITY R. [ eld 1'ey, Mana or Nuclear Licensing and Regulatory Affairs p) cc: Sce page 2
\\ \\
An Equal Opportunity Employer
r 9
' U.S. Nuclear Regulatory Coumission
/{ h gg Rnclosure cc (Rnclosure):
Ms. S. C. Black, Assistant Director for Projects TVA Projects Division U.S. Nuclear Regulatory Commisalon One Whita Filnt, North 11555 Rockville Pike Rockville, Maryland 20852 Mr. F. R. McCoy, Assistant Director for Inspection Programs TVA Projects Division U.S. Nuclear Regulatory Commission Region II 101 Karietta Street, NW, Suite 2900 Atlanta, Coorgia 30323 Browns Ferry Resident Inspector Browns Ferry Nuclear Plant Route 12, P.O. Box 637 Athens, Alabama 35611
r s
,t C
ENCLOSURE Browns Ferry Nucl iar Plant Probabilistic Risk Assessment (BFN PRA)
Summary Report
~ ~ ~
As requested by the NRC staff (reference NRC letter dated May 27, 1988), this Summary Report provides:
- TVA's rationale for concluding that the latest BFN PRA will conservatively. reflect the configuration of BFN Unit 2 at the time of restart and
- Summary of the changes made between the January 1986 version of the BFN PRA reviewed by the NRC staff in its October 1, 1987 letter and the September 1987 version.
The purpose of this report is to provide aeditional information to support a conclusion that'BFN is not an outlier with respect to severe accident characteristics when compared to plants of similar a
type and vintage.
e
7 TABLE OF CONTENTS P.!ULe 1.0_. EXECUTIVE
SUMMARY
AND CONCLUSIONS,
1-1 4
2.0 INTRODUCTION
2-1 3.0 MAINTAINING PLANT CONFIGURATION IN THE PRA 3-1 3.1 Confirmation of Design Baseline.
3-1 3.2 Process for Evaluation of Future 3-1 Design Changes 4.0
SUMMARY
OF BFN PRA CHANGES 4-1 4.1 Model Rafinements 4-1 4.2 Changes in Analyses 4-2 4.3 Assessment at Operator Actions 4-2 5.0 EFFECT OF CHANG'ES 5-1 4
5.1 Component Failure Data and 5-1 Initiating Event Frequencies 5-2 5.2 System Unavailability Impacts 5.3 Core Melt Frequency Profile 5-2 6.0 BFN SEVERE ACCIDENT CHARACTERISTICS
~
6-1 LIST OF TABLES Tab 1_e Title 4-1 Status of Potential Refinements Identified T-1 in Table 6.6-3 of January 1986 BFN PRA 4-2 Additional Refinements Made in the PRA T-9 4-3 Overview of Changes in Analyses T-12 5-1 Impact of Changes on Component Failure T-13 Data and Initiating Event Frequencies 5-2 Impact of Changes on System Unavailabilities T-17 5-3 Changes to System Importance T-20 5-4 Importance of Internal Initiating Events T-23 6-1 Plant Damago State Frequencies T-25 6-2 Dominant Sequences T-27
-i-
r 1.0 EXECUTIVE
SUMMARY
AND CONCLUSIONS The BFN FRA makes use of realistic mode).ing assumptions and plant specific data to the greatest extent practical. The level of detail of the analyses is consistent with that found in other PRAs performed by the industry. The total core damage frequency estimated for BFN Unit 2 is 4.7x10~*/ year, within the range typical for similar BWR plants. No sequence of events contributes greater than 5% to the. total core damage frequency. No new generic safety issues have been identified by the analyses.
The September 1987 version of the BFN PPA conservatively reflects the configuration of BFN Unit 2 based on completion of a review of plant drawings and implementation of a review process to evaluate future changes to the plant design.
Changes made between the Jacuary 1986 PRA and the September 1987 version are primarily due to:
- Refinement of the models used,
- Analyses initially done as scoping completed in detail.
- Addition of more plant specific data, and
- Changes in assessment of operator actions.
Page 1-1
r
2.0 INTRODUCTION
The BFN PAA is a full scope probabilistic risk assessment of the Browns Ferry Nuclear Plant which includes:
- Core damage accident sequence analyses.
- Containment (Mark I) response analyses, and
- Site specific consequence analyses.
Initiators: considered include transients, loss of coolant acaidents (LOCAs), and events such as earthquakes, fires, internal and external flooding, high wind, aircraft impact and turbine missiles. Systems and sequence analyses were performed in detail.
Initiating events that could result in core damage occurrence were analyzed and systems needed to mitigate these events b~;e modeled to estimate the core damage frequency.
Development of a PRA is an iterative process. During the continual review and improvement activities inherent to the PRA process, refinements to the system models, hardware data and initiating event frequ'ncies are identified. Refinements also e
result from the ongoing evaluation of changes to the plant configuration and procedures as discussed in Section 3.0.
The major changes incorporated into the BFN PRA September 1987 version are discussed in greater detail in Sections 4.0 and 5.0.
Page 2-1
3.0 MAINTAINING PLANT C0FFIGURATION IN THE PRA A two-phase review program has been implemented to assure that the PRA will conservatively reflect the BFN plant configuration at restart. The first phase consisted of a drawing review to establish a design reference base from which the continuing change monitoring and review program of the second phase builds.
3.1 confirmation of Design A review of BFN drawings (Phase 1) was performed to determine if there were significant differences between the PRA model and the current plant configuration.
Several sets of information were found which provided overlapping information. An evaluation was made to determine and select the most current set of information.
An information hierarchy was established for the BFN Unit 2 drawings as follows:
1.
As-Constructed and Verified 2.
As-Constructed 3.
As-Designed A drawing "freeze" date of May 15, 1988 was selected. A list of drawings referenced in previous versions of the PRA was then prepared. The latest revisions of these drawings were obtained, filed, and declared "current" to May 15, 1988. Modifications to the plant and revisions to these drawings after this date are addressed in Phase 2, the change monitoring and review program.
~
The referenced drawing list was sorted by PRA system and distributed to the reliability engineers. Each engineer reviewed these drawings to identify and evaluate the
. significance of the differences between the PRA models and the current revision of the drawings. These differences were documented on a "Drawing Review Form." No changes were identified in Phase 1 which significantly influenced the results of the PRA.
3.2 Process for Evaluation of Future Design Changes Phase 1 documented the accuracy of the PRA models up to the plant configuration freeze date of May 15, 1988.
The change monitoring and review program (Phase 2) provides an overview of the changes in the plant configuration after the freeze date to determine any effect on the PRA models and conclusions. This program is not intended to continuously revise the PRA models to reflect the actual plant configuration.
Instead, the program ensures that, between periodic upostes, the conclusions reached as a result of using PRA analysis remair. accurate.
Page 3-1
Chang 2s in tha picnt which involva physical modifiestions are documented by drawing revisions and through engineering change notices (ECNs). Revised drawings are distributed to and reviewed by reliability engineers for potential impact on BFN PRA conclusions.
In a manner similar to the revised drawing review, the ECNs designated as "drawing complete" or "closed" are collected and processed through the change monitoring and review program.
The revised drawings and the ECNs designated as "drawing complete" or "closed" are combined into Change Evaluation Packages. Each package is distributed to the reliability engineer responsible for the affected system for evaluation of the significance of the plant changes on the model and the PRA conclusions. The engineer documents the evaluation using the "ECN/ Drawing Evaluation Form."
The entire Change Monitoring Package containing the ECNs, a list of revised drawings, and ECN/ Drawing Evaluation forms are retained.
Essentially, three Phase 2 evaluation conclusions related to the impact of plant changes on the PRA are:
'No impact; the change does not involve modeled equipment and does not need to be specifically
- modeled,
'The change involves modeled equipment and decreases or insignificant 1y increases risk which results in conservatism in the PRA conclusions, or 1
'The change involves modeled equipment and significantly increases risk which results in less conservatism in the PRA conclusions.
Changes involving modeled components will be appropriately incorporated in periodic revisions to the PRA. For the case where risk is significantly increased, the BFN project engineer will be notified.
1 Page 3-2
4.0
SUMMARY
OF BFN PRA CHANGES Changes made in the BFN PRA since the January 1986 version are discussed in the following sections. These changes come from the following sources:
'The potential refinement actions listed in Table 6.6-3 nf the January 1986 PRA, and
' Modifications identified during the iterative PRA analysis process and reviews by various organizations.
4.1 Model Refinements Many conservatisms and assumptions in the January 1986 version were identified which, if more rigorously evaluated, could result in more realistic models and would tend to decrease the calculated overall core damage frequency.
These potential PRA refinement actions, identified in Table 6.6-3 of the January 1986 version, and the extent of their incorporation into the September 1987 version are shown in Table 4-1.
Of the 56 potential refinements identified in Table 4-1, 23 have been fully or partially incorporated 31 have been i
deferred, and 2 are deemed not possible at this time. Of the 31 refinements deferred, 19 require more detailed recovery analysis 3 require more data assessment and 9 require a more realistic definition of success criteria.
The 2 refinements determined not possible to incorporate at this time were so designated primarily due to their complexity. The refinements not included are not expected to significantly impact the estimated core melt frequency for BFN.
Additional refinements were made to system models, hardware data, and initiating event frequency as shown in Table 4-2.
Some of the refinements listed in Table 4-2 resulted from the incorporation of changes in plant design or procedures. The basis for the system models was l
changed from the as-designed drawings to the as-constructed drawings, where available. System models were transferred from the Discrete Probability Dictribution (DPD") computer code to the RISKMAN" computer code. Other changes resulted from reviews of the i
PRA by various organizations and personnel internal and external to TVA. This resulted in the systems being reviewed from a new perspective, generating additional refinement actions.
l l
l I
Page 4-1 l
i
F 4.2 Changes in Anslyssa Various analyses of potential risk significant events have been completed or revised for the September 1987 version of the BFN PRA. Table 4-3 provides an overview of the status of completion of these analyses. Table 4-3 identifies which analyses were done in detail or as scoping analyses and which analysos were revised. As a minimum, a scoping analysis has been performed for these events except the loss of electrical boards below the 4KV level. For the loss of electrical boards below 44V level, detailed dependency matrices were developed and included in the September 1987 version of the PRA.
4.3 Assessment of Operator Actions Operator actions were assessed in the January 1986 version of the BFN PRA by assessment of the desired action, comparison with the values of similar actions in other PRAs, and engineering judgment. Assessment of the desired action involved:
- consideration of the timing of the scenario progression.
- ease of diagnosing the event (i.e., the availability of unambiguous indications and operator feedback),
- the relative stress level experienced by the
- operator,
- review of applicable plant procedures, and
- review of the scenarios with the plant operators.
Coupled errors, which occur when one operator error influences the likelihood that a second error will occur, were also included. Since coupled errors con introduce dependencies between top events
- and since operator actions were not considered as separate top events, hardware and operator contributions to top event unavailability were combined for use in the event tree.
The resulting equivalent operator error rate represents the direct human error induced failure, allowing recognition that several opportunities may exist for operator intervention and that one action may affect more than one top event.
- A top event is a system, component or operator function that determines the progression of plant responses to an initiating event.
Page 4-2 m-m
l Ths most significant chtngss to assessment of oparctor action between the January 1986 version and the September j
1987 version of the BFN PRA are related to the evaluation of the probability that high pressure makeup failure due to operator action or inaction will lead to subsequent operator error in failure to manually depressurize the reactor vessel. Event-oriented emergency operating procedures used previously were replaced by symptom-based procedures, which significantly increase the likelihood of successful depressurization and decrease the extent of coupling of this action from manual high pressure coolant injection / reactor core isolation cooling (HPCI/RCIC) control.
Symptom-based procedurea also increase attention on torus water temperature, reducing the uncoupled operator action contribution by a factor of 10 for sequences reetiring establishment of torus cooling several hours into a scenario. The value for the latter action is conservative when compared with the results of a detailed formal analysis contained in a recent BWR PRA.
A more detailed description of the human factor analysis method used is provided in Appendix A.
4 Page 4 3 i
r
5.0- EFFECT OF CHANGES The changes made between the January 1986 PRA and the September 1987 version were described in Section 4.0.
The effects of these changes are sunnarized in the foll.owing sections.
5.1 Component Failure Data and Initiating Event Frequencies For the January 1986 version, plant specific data had been collected from the plant startup to September 1980. As identified in Table 4-2, the collection period for plant specific data has been extended from September 1980 to the end of.1985 for the September 1987 version.
This resulted int
- A decrease in some hardware failure data values by as much as a factor of five,
' 'An increase in some hardware failure data values by as much as a factor of four, and
'No significant change in some values.
The extended col'lection period for plant specific data resulted in the data values being more representative of actual Browns Ferry operating history. The failure data used for the January 1986 version and for the September 1987 version are shown in Table 5-1.
A second part of Table 5-1 lists the frequencies for the internal initiating events used for the January 1986 and September 1987 versions of the PRA.
As discussed in Section 4.1, the categorization of plant specific events was reviewed and revised.
For example, the loss of feedwater event was divided into three subcategories. This resulted in a decrease in the importance of loss of feedwater events by allowing recovery of the feedwater system to occur. Also, the estimated plant capacity factor used to convert initiating event frequencios to calendar years in the January 1986 version was changed to the actual average plant capacity factor. These changes resulted in an overall decrease in initiating event frequencies by a factor of approximately two.
i Page 5-1
{
i l
1
- -. ~ ~
~,-.c
-n---c,
-e-E
E 5.2 System Unnvtilsbility Irpreto Table 5-2 lists the system unavailabilities used to quantify the event trees for the January 1986 and September 1987 versions of the PRA. The unavailabilities listed are for all Common Actuation Sensors (CAS)* signals available and all electric power available (referred to as CAS state 1 and electric power state 16). The system unavailabilities for other combinations of CAS and electric power states are not necessarily the same as '
those listed in Table 5-2.
However,-the system unavailabilities for CAS state 1 and electric power state 16 are used since they give the best representation of the system unavailability (no electric power or CAS dependent failuren).
The system unavailabilities differ for the January 1986 and September 1987 versions of the PRA due to the incorporation of the refinement actions listed in Tables 4-1 and 4-2.
In general, the unavailabilities have decreased from the values used in the January 1986 version.
In a few cases, however, the uravailabilities increased. For, example, the use of plant specific data resulted in an increased unavailability for relief valves reseating (top event Ml) and the use of updated design information resulted in an increase in the unavailability for recirculation pump trip (top event RP).
5.3 Core Melt Frequency (CMF) Profile For each version of the PRA, the top 100 core damage scenarios were carefully reviewed.
For the purposes of this summary report, the contribution of an event category (or system top event) to the CMF is interpreted as the "importance" of that event category. The importance of a particular event (or system top event) is defined by the sum of the frequencies el the individual top 100 sequences which contain that particular event (or system top event). Care was taken to assure that system top event importance does not include the sequences in which support system failures guarantee top ? vent failure. This was done to prevent obscuring the tru importance of the system itself. Support system importance is accounted for separately in Table 5-2.
- CAS is a BFN PRA term used to represent those sensors which provido a common accident actuation signal to multiple components and systems.
I Page 5-2
Systca icval importcnce is tha sum of tha importcncs of the top events of that system.
Because the event model differentiates between plant damage states in core melt scenarins (e.g., the availability of debris bed cooling is determined), the relative importance of the residual heat removal (RHR), core spray and condensate systems may be inflated.
In many sequences, the success or failure of these systems does not determine whether or not a core melt occurs, rather they cnly affect which plant damage state is entered.
Tables 5-3 and 5-4 summarize the relative importance of the systems and internal initiating events, respectively.
l It should be kept in mind that the importance measures are I
relative. That is, the measure gives the importance of a particular initiating event or system top event in relation to the other events for that version of the PRA only.
If the total CMF were the same for the two versions, then direct comparison of the relative importance values for the two version would be appropriate. However, since the CMF decreased by a factor of 10. direct comparison of the importance of an event from one version of the PRA to another could be misleading. This is because the event's relative importance may have increased while its absolute importance has decreased due to the overall decrease in CMF. For this reason, it is more appropriate to compare changes in the core melt frequency profile for each version, characterized by a collective view of the importance measures.
Table 5-3 indicates the overall relative importance of the RHR system decreased while the importance of the torus cooling mode of RHR increased. This resulted from refinement actions which decreased the requirements on the low pressure coolant injection (LPCI) mode of RHR but increased the requirements for torus cooling. The apparent large increase in the importance of HPCI/RCIC can be attributed to the redefinition of top event HP (automatic start of HPCI or RCIC). A part of thio increase is also due to the inclusion of the control rod drive (CRD) system along with HPCI/RCIC in top event HP, which allows a more accurate assessment of the capability l
of the CRD hydraulic system (CRDHS) to provide high pressure injection. The refinement actions also increased i
the relative importance of emergency equipment cooling water (EECW) and the balance of plant systems. The Importance of the reactor protection system (RPS),
I recirculation pump trip and the control air system decreased.
l l
Page 5-3
I T;blo 5-4 pervid:s en indicaticn cf tha impact of tha refinement actions on the importance of the initiating event categories for the January 1986 and the September 1987 versions. The refinement actions tended to increase the relative importance of transients that resulted in loss of the secondary plant systems (e.g., main steam isolation valve (MSIV) closure, loss of condenser vacuum, and loss of offsite power). At the same time, the refinements decreased the relative importance of events involving loss of control air, stuck open relief valves, and anticipated transient without scram (ATWS). This is expected since these events dominated the January 1986 core melt profile and the refinements were prioritized on the basis of impact on the dominant contributors. As indicated in Table 5-4, the core melt frequency for BFN is generally due to transients involving balance of plant systems.
Tables 5-3 and 5-4, which ifst the importance of systems and internal' initiating events, respectively, also show that the calculated core melt frequency from the internal events has been reduced by a factor of ten through incorporation of changes between the January 1986 and the September 1987 versions.
4 I
r Page 5-4 L
6.0 BFN SEVERE ACCIDENT CHARACTERTSTICS The January 1986 version of the BFN PRA was known to overpredict the actual core melt-frequency of the plant.
Changes implemented in the BFN PRA have,been focused to remove unnecessary conservatisms, incorporate plant changes in the PRA models, and refine the analyses to more closely reflect the actual plant' configuration.
Information gained from changes in plant design and procedures, increasing operating history, advances in analysis techniques, and PRA reviews resulted in refinement and improvement of the PRA analysis.
The September.1987 version of the BFN PRA makes use of realistic modeling assumptions and plant specific data to the greatest extent practical, with a level of detail of the analyses consistent with that found in other PRAs performed in the industry.
An overall effect of the refinements made has been to reduce the estimated CMF. Table 6-1 lists the individual plant damage state frequencies and the total CMF for the January 1986 and the September 1987 versions of the PRA. The current core melt 4.7x10-*y due to internal initiating events for BFN is frequenc
/ year. The plant damage state frequencies are composed of individual event sequences grouped by common 6Efcets on the state of the plant.
The dominant sequences for the September 1987 version of the BFN PRA are identified in Table 6-2.
As indicated in Table 6-2, the core melt frequency for BFN is generally due to transients involving balance of plant systems. No sequence of initiating events contributes greater than 5% to the total core melt frequency.
Therefore, it may be concluded that the September 1987 version of the BFN PRA is an accurate representation of the plant's risk and that Browns Ferry is not an outlier plant in relation to severo accident characteristics for plants of similar type and vintage.
Page 6-1
e e
APPENDIX A ASSESSMENT OF OPERATOR ACTIONS METHODOLOGY Care was taken in acknowledging the possibility of "coupled" operator errors occurring. Coupled errors occur when one error influences the likelihood that a second error occurs. Such errors may introduce dependencies between several top events.
Such coupled errors were included in the BFN PRA.
Because operator actions were not identified as separate top events in the PRA, it was necessary to combine hardware and operator contributions to top event unavailability for use in the event tree.
An example of how this coupling was handled is the assessment of operator errors involving manual control of the injection systems and manual depressurization. High pressure vessel makeup by the high pressure coolant injection (HPCI) or reactor core isolation cooling (RCIC) system in transient events is defined as top event HP.
Manual depressurization is defined as top event V2.
For most transients, the value for top event HP that is used in the event tree quantification,is given by HPr = HPn + HPo - HPn HPo The H and 0 subscripts refer to the hardware and operator contributors to unavailability, respectively.
Both HPCI and RCIC receive an automatic start signal on level 2 and a trip signal on level 8.
Following a high level trip, HPCI has the potential to continue to automatically cycle; hardware failures (e.g., HPCI pump fails to start on demand) are possible during cycling. A HPCI/RCIC cycling model was developed which considers the operator as an integral part of the process. Critical timing, alarms and instrument information are a function of initial vessel refill rate (e.g., whether HPCI is available) and the location within the cycle. The resulting value for HPo is interpreted as an equivalent operator error rate and represents the direct human error induced failure of HPCI/RCIC recognizing that several opportunities exist for the operator to take control of HPCI/RCIC flow.
The expression for the closely related top event V2 for the case where HP has failed is given by:
V2=((1-HPa)HPo(V2n + (1-V2n)V2o*] +
HPn ((1-V2n)V2a + V2n))/HPr The parameter V2o* is interproted as the probability that the operator f ails to manually depressurize the vessel af ter high pressure makeup (HPCI and RCIC) fails due to operator action or inaction.
In this manner, the potential coupling of operator errors between top events HP and V2 is explicitly accounted for in the quantification of the event trees.
a TABLE 4-1 Page I of 8 Status of Potential Refinements identified in Table 6.6-3 of January 1986 PRA Designator Refinement (Il from Table 6.6-3 Actions Conuent A
Modify ATWS model to reflect ORNUSASA Partial High Pressure makeup with RCIC and CRDHS results on high pressure injection added. The ATWS model is still conservative.
sucess criteria 8
Refine high pressure makeup used in Partial Iterative refinements made in the operator ATWS model response model.
C Refine IFS Top event definitions to incorporated Added second input trip signal to RPS top Include additional signals events.
D Include ATWS sequences for 1-3 Incorporated in the original event model, sequences went stuck Open Relief Valves (SORVs) directly to core melt for I-3 SORVs and with scrare failure RPS failure.
E include manual actions to start incorporated Replaced top event R1 with R7. Top event lGR purys for torus cooling (RI) and R7 includes manual start. RI does not.
refine hurnan error model for I-3 SORVs F
Includes manual start of RHR (and incorporated Added manuel start to top event RD.
refine human error model) for multionit events (Also consider.
I IUR pwp and i RIR Service Water (RIESW) pump with 1/3 flow)
G Reconsider model of doisel generator Deferred Involved changing diesel generator failure failure during operation rate to time dependent failure. Inportance of this refinement action deemed small.
EECW doeinates station blackout frequency.
T-1 e
a
TABLE 4-1 Page 2 of 8 Status of Potential h fine m ts identified in Table 6.6-3 of January 19R6 PRA Designator RefInoment(ll frrn Table 6.6-3 Actions Consnent H
Include recovery frae Standby Liquid Deferred Inportance of this refinement action Control (SLC) SI (ATWS) decreased with the irrmation of other ATWS related refinements.
I Manually open shutdown cooling valves incorporated Based on actual plant experience.
J Recover air Je.g., use bottled air)
Deferred Inportance of this refinement action decreased with the incorporation of other control air system refinannnts.
1 K
Include CRDHS and multiunit equivalent Not Possible Refinement action complex to incorporate; on of manual start of IUR ptmps for LOSP, no power to OtDHS pumps. Loss of I
torus cooling (R7) for loss of control Control Air Model conservatively envelops air and loss of offsite power events loss of Raw Cooling Water (RCW) and loss of (LOSP)
Reactor BuildirIg Closed Cooling Water (ItBCCW).
L Consider transient recovery (i.e.,
Deferred Likelihood of successful retur. to power not return to power) for successfoi evaluated. No experience base available.
bIowdown cases M
Include CRDHS for sequences with suc-Not Possible Complex to incorporate due to CRDHS dogmw cessful manual depressurization; low doncles. Would have to be done on a pressure makeup; and, containnent sequence by sequence basis.
cooling N
Recategorize relief valve da*a Deferred Analysis required to differentiate plant (transient /l-3 SORVs) response for differing nunber of relief valves stuck open.
T-2 t
I
o TABLE 4-1 Page 3 of 8 Status of Potential Refinements identified in Table 6.6-3 of January 1986 PRA Designator Refinement (I) 1 rom Table 6.6-3 Actions Comment O
Restructure transient events to reflect Deferred Refinement action complex to incorporate.
nuser of relief valves actually lif ted W uld require detailed analysis to identify nud er of valwas Iiffing for difforent sequences.
P Manual action to start IFCI and EECW Incorporated Action Q is a parallel path to action P.
(H4, EE) buy time to manually open RER injection valves (e.g., transient /
loss of coolant accident (LOCAs)
(See QI Q
Manual action to start IfCI and EECW Deferred Incorporation of Action P reduced benefit (H4 EE); include CRDHS and manual of this action.
start of RIR for torus cooling (e.g., transient /LOCAs) (See P1 R
Reenalyze IE data (I SORV versus 2-3 Deferred Existing data cons'idered conservative.
SORVs; exclude electromatic and 3-stage Two-stage Target Rock data was used in valves) ttw ~,afety Relief Valve (SRV) systems analysis in the January 1986 version; initiating event prior population data contains data from all US BWRs.
S Manual start of ifCl and EECW (H3, E2);
incorporated Action T is a parallel path to action S.
buy time to open RER injection valves (small LOCA) (see T1 T
Manual start of ifCI and EECW (H), E2);
Deferred incorporation of Action S reduced benefit of and manual start of RIR for torus this action.
cooling (small LOCA) (see SI U
Manual start of IfCl and EECW (IP, EE);
Incorporated Action V is a parallel path to action U.
buy time to open RIR injection valves (transients) (see V1 I
T-3
- ~
O g
TABLE 4-1 Page 4 of 8 Status of Potential Refinements identified in Table 6.6-3 of Januar= 1986 PRA Designator Refinement (I) 3 from Table 6.6.3 Actions e-t V
Manual start of IFCI and EECW (HP, EE);
Deferred Incorporation of Action U reduced benefit of and RIR pops for torus cooling in model ttils action.
transients (see UI W
Resort loss of Feedwater (LOF) IE data incorporated Loss of feedwater broken up into throe categorics; total recoverable, total nonrecoverable, and partial loss.
Previously, any LOF (complete or partial) was assumed to be total loss.
X Replace f5R top RB witti R7 (re-examine Incorporated Originally, top IB required auto-start human error model; small LOCA) of IUR pumps and manual alignment to torus cooling. Replaced with R7, manual start of torus cooling.
Y Reconsider operator action portion of incorporated Refined operator error pcction of le top manual depressurization for cases event model.
other than IFCI and RCIC Initial failure (see GG, AF)
Z include transient recovery af ter I.
Deferred Requires analysis to determine likelihood of IFCI/RCIC vessel fill after pressure successful recovery.
regulator fails closed AA Include transient recovery after I Deferred See Z.
IFCl/RCIC vessel fill after MSlV closure BB Include transient recovery after i Deferred See Z.
IFCl/RCIC vessel fill after loss of feedwater T-4 t
i
\\
TABl.E 4-1 Page 5 of 8 Status of Potential Refinements identified in Table 6.6-3 of January 1986 PRA Designator Refinement (I) f rcum Table 6.6-3 Actions,
Comument CC Reconsider MSIV closure initiating event Deferred Differentiation of Intial power levels deemed data (single valve closure events below a to not be beneficial compared with complexity certain power level would not result in to incorporate refinement.
closure of all valves )
DD include transient recovery af*er i Deferred "
See Z.
IPCl/RCIC vessel fill after loss of condenser vacuum EE Reanalyze :nitial control of feedwater Deferred Current model is conservative.
FF Reconsider model coupling operator action incorporated Revised model coupling human error.
Involved with foedwaier, HPCI, RCIC Control and manual depressurization GG Reconsider conditional human error Deferred lacorporation of action Y decreased associated with manual depressurization importance of this action, j
for foodwater rampuo M -
tel Recover 480-volt board IA Incorporated Incorporated on sequence specific basis.
Il Recover 4-kV shutdown board fran relay incorporated Incorporated on sequence specific basis.
test JJ Recover 480-volt board IA before lack Deferred Analysis required to determine time available of room cooling faiis pumps tiII IGR pumps falI.
KK Recover 4-kV shutdown board IA before Deferred See JJ.
lack of roose cooling fails punps af ter LOSP T-5 1
e
y
+
TABLE 4-1 Page 6 of 8 Status of Potential Refinements identified in Table 6.6-3 of January 1986 PRA Refinement (II Designator from Table 6.6-3 Actions P-t LL Remove relief valve reseating top (MI)
Incorporated Number of stuck open relief valves has no from ATWS model inpoet on plant re:ponse to ATWS based on ORNL analysis.
Im Reconsider model coupling operator incorporated Recent model accounts for availability of action involved with IFCI and RCIC.
new symptom based Emergency Operating control and marreak aepressurization Procedures (EOP's).
coupling model Ici Recover 480-volt board IA before t Deferred See JJ.
of room cooling falls pm af ter t0SP 00 Recover 4-kV shutdown board before Deferred See JJ.
lack of room cooling fails pumps after LOSP PP Recover power after i fill by IFCl/RCIC Deferred Analysis required to determine bolloff time after LOSP from level 8 rather than level 2.
QQ Reconsider diesel generator comon Partial Updated conson cause parameters utilized in cause model; incorporate single September 1987 version.
and double diesel ger. orator recovery into electric power system model RR Reformulatr,model to take credit for Deferred Analysis required to verify success criteria.
I core spray pump supplying makeup after successful operation of ifCI, RCIC; include top event RD and 480-V board recovery after I.OSP T-6 I
TABLE A-l Page 7 of 8 Status of Potential As.finements idos.tified in Table 6.6-3 of January 1986 PRA Designator def.* m t(3?
fre T.::le 6.6-3 Actions conomat Like action IWt oncept IHt flou is split Deferred Ser M.
to provide both long-term mekaup as voll as torus cocling after LOSP
T Use eulsting IHL pisaps to reflood; cycle Deferred See IWt.
as necessary, use I core spray pump to maint,In levet af ter LDSP include multiple (up to 3) diesel Partial see QQ Above generator recovery reconsider consuon cause model VV Inct 4 ::asitiple t p to 4) diesel P rtial See QQ Above generator recnvery; reconsider common cauw model W
Re-examine MSIV closure (GI); develop data Inc~cporated initial versions of PRA used generic valve for M58V4 data for MSIVs. September 1987 version used MSIV specific data.
XX Treat M5tV closure failure llPm large I* %rred luportance of this refinement action LOCA; use lou pressure. energency decreased with the incorporation of cooling system (LPECS); must consider action W.
j long-term Inven.ory makeup YY Include manual / alterative actions to Deferred importance of this refinement action close MSIVs/ Bypass valves and turbine decreased with the incorporation of control valves (TCV) or turbine action W.
stop
'I m (TSVs) to terminate tre AB Recover 49t or use other unit IHt to Deferred Requires multi >.mit modeling to determine cool torus availability of other unit IHL pumps.
I T-7 v
m.
g
\\
TABLE 4-1 Page 8 of C Status of Potential Hefinguents identified in 1
Table 6.6-3 of January 1996 PHA Lesignator h finemer.t(I) from Table 6.6-3 Actions r-+
AC Heenalyze operato action associated Deferred Cannot be incorporated until action 0 and R with I PCI with timing considerations are incorporated.
specific to I or 24 SORis AD Eliminato ItR for containment spray (SP)
Incorporated Requirement for containment spray for large requirement break events deleted based on ORNL analysis.
l AE Include IDIR (IS) and 2 loops.-f core Caferred Requires analysis to verify success critoria.
j spray as a success sequence fo-large M
i 1
I AF Reconsider operator action associated Deferred See GG.
J i
with manual depressurization fc.?
feedwater ranpup sequences Note (1):
These acties represent a ceabination of refinements perfonned in the Iteritive revision process talwoon the January 1986 and me Septenber 1967 versions of the BFN FRA.
Note (2):
Incorporated: %"asnent Action incorporated Partial:
Refinement Action Partially incorporated 1
Deferred:
h firmaant Action Deferred Not Possibir,r Refinement Action would require considerable modeling effort or not possible l
l T-8 l
i TABLE 4-2 P go 1 of 3 ADDITIONAL REFINEMENTS MADE IN THL PRA I
I.
SYSTEM MODEkfilANSES Sygien Additional _ Refinements (l)
Commor Actuation Eliminated common cause gamma codell incorporated Sensors Analog Trip System (ATS) modification Residual Heat Removal Revired ma!.ntenance model to reflect actual Service Water EECW/RHREW pump maintenance practices Raw Cooling Water Removed requirement for Auxiliary Raw Cooling Water System Emergency Equipment Included recovery from logic system survol11ance Cooling Water test; manual start of pumps included Recirculation Pump Upgraded to reflect latest design information and ATS Trip Condensate Cycling /startup bypass valve model modified to more a
correctly represent actual case Drywell control Interconnection with plant air includ J, minor Air corrections to equations Electric power Recovery Expanded for Offsite Power and 480v ac System Shutdown Boards 1A and 1Bt models for 250v de RMOV board 1B and Rattery Board 3 reviced for LOSP events.
Reactor Protection Second input signals added for top events T3 and System T6 to reduce dependence on input signals; incorporated ATS modification Residual Heat Minor corrections to equations Removal Primary Containment Incorporated ATS modification; model requantified Isolati' utilizing MSIV specific data T-9 E
~
l TABLE 4-2 P ga 2 of 3 ADDITIONAL REFINEMEN'S MADE IN THE PRA II.
EVENI_MQDEL CHANGES
/.dditionaLRefdncmenta(1)
Transient Events Models revised reflect new E0Isl ADS replaced by involving ADS manual depressurization Transient Events Expanded use of early manual initiation of RHR and involving EPS 1 core sprsy for seqtences in CAS state 1 or 2 and with early depressurization Transient Events Ref;'.nement of HPCI/RCIC cycling modell CRD included where approprikte Loss of feedwater Models developed for three subcategories of loss of feedwater MSIV Closuret Loss of Firures of separate eveat traes without recovery Condenser Vacuumt deleted iron document Pressure Regulator Failure-closedt loss of Plant Alr; lo.1s of feedwater III. COME0hENLEAILVRE_ DATA _CILMGES Cn9ponent AdditionaLRefinemeaAa HPCI pump Data base updated to relleet system modifications RCIC pump Collection period extended for plant specific data Core Spray pumps Collection period extended for plant spec!.fic data RHR pumps Collection period extended for plant specific data Standby Liquid Collection period extended for plant specific cata Control pumps RHRSW/EECW pumps Collection period extended for plant specific data Feedwater pumps Collection period extended for plant specific data Condensate pumps Collection period ext (ndeo for plant specific data T-10 c.
TABLE 4-2 P ga 3 of 3 ADDITIONAL REFINEMENTS MADE IN THE PRA I
III. COMEONELEAILURE DATA CHANGIS (Continued) 4 Componsni Additional Refinemenia Condensate Booster pumps Collection period extended for plant specific data Motor Operated valves Collection period e'xtended for plant specific data
- ~
Manual v119ss Data for failure mode "failure to open on demand" added Main Steam Relief Valves Collection period extended for plant specific data Main Steam Isolation Collection period extended for plant specific data; Valves Data for failure mode "failure to close on demand" added RHR Her.t Exchangars Collection period extended for plant specific data RHR room cooler Collection period extended for plant specific data CS room cooler Collection period extended for plant specific data Air compressors Collection period extended for plant specific data Control Air dryers Plant data reviewed and refined Air Relief Valves Data for 1/2" to 2" spring actuated valves "leaking sticking open or lifting prematurely" added
- 17. 1HIU AUNG_EV MIJEQUENCLDAIA_ CHANGE S All Internal Initiating (1) Collection period extended for plant specific Event Uategories data.
(2) Categorization of all plant specific events reviewed and revised as necessary.
(3) Capacity factor used to express frequency in terms of calendar years revised.
Note (1): These actions represent a combination of refinements performed between the January 1986 version and the September 1987 version of the BFN PRA.
T-11 t
s TABLE 4-3 OVERVIEW OF CHANGES IN ANALYSES January September 1986 1987 Version Version Internal Events as Described in Section 6.4 of the PRA A
AR Other Events as Described in Section 6.5 of the PRA Loss of Electrical Boards N
N Loss of HVAC N
S Common lustrument Tap Considerations N
A Interfacing LOCAs N
A Multiple Unit Interactions N
S Loss of the Condensate Storage Tank N
S Torus Rupture S
A MSRV Taf.1 pipe Vacuum Breaker Stuck Open N
S Scram Discharge Volume Break N
A Loss of Decay Heat Removal N
S Loss of Reactor Building Closed Cooling Water N
S Purging During Operation N
S Inadvertent Fire Suppression System Operation N
A Loss of Recirculation Pump Seal Cooling N
S Loss of Flow Through Traveling Screens N
S Common Accident Signal Consideration N
S Radiological Releases by Means Other Than the Core N
A External Events as Described in Section 8 and Appendix E of the PRA Earthquake A
AR Extreme Winds and Tornado A
A Aircraft Impacts A
A Fires A
S F1 coding (External A
A Flooding (Internal)
A AR Transportation Accidents S
S Toxic Gas Release S
S Containment Response Model S
S Offsite Consequence Analysis S
S At Analysis Performed art Analysis Revised S.
Scoping Analysis Performed Nt Analysis Not Performed
s TABLE 5-1 P:gs ! cf 4 IMPACT OF CHANGES ON COMPONENT FAILURE DATE INITIATING EVENT FREQUENCIES I.
COMPONENT FAILURE DAIA All values shown are the mean of a distribution and given as occurences per demand or per hour, as noted.
Janua'ry 1986 September 1987 Component /Fa11ure Ptoda Vera1Q0.,_
Veraion___
HECLpump failure to start on demand 6.62x10-2 2.41x10-2 failure during operation (per hour) 3.95x10-4 3.87x10-4 BCIC_ pump failure to start on demand 6.06x10-2 3.98x10-2 failure during operation (per hour) 3.95x10-4 3.87x1 -4 Core _ Spray _pumpa failure to start on demand 1.26x10-3 9.72x10-4 failure during operation (per hour) 2.62x10-5 2.62x10-5 RHILpumps failure to start on demand 1.6x10-3 1.83x10-3 failure during operation (per hour) 1.76x10-5 1,41x10-5 Standby _ Liquid..ContrDLaumps failure to start on demand 2.27x10-3 1.85x10-3 failure during operation (per hour) 2.62x10-5 2.61x10-5 RHRSW/IECW_pumpa failure to start on demand 3.44x10"3 6.25x10-4 failure duries operation (per hour) 6.47x10-5 4.69x10-5 Esadwater_pumpa failure during operation (per hour) 2.85x10-5 4.99x10-5 Condensate _ peps failure during operation (per hour) 1.97x10-5 1.55x10-5 Condensate _ Booster _fumps tailure during operation (per hour) 2.64x10-S 2.21x10-5 tietoI_0perated.Jalrea failure to operate on demand 3.28x10-3 2.45x10-3 4
.s
~.
TABI.E 5-1 Pcgo 2 of 4 IMPACT OF CHANGES ON COMPONENT FAII,URE DATE INITIATING EVENT FREQUENCIES January 1986 September 1987 Component / Failure Mods Version Version Manual _VA1xts failure to open on demand 2.13x10-4 Main _ Steam _Relitf_Valrea failure to open on demand 3.63x10-3 4.0$x10-3 failure to reseat on demand 2.63x10-3 5.22x10-3 Main _Staam_ Isolation _Valysa failure to open on demand 2.80x10-4 1.93x10-4
)
failure to close on demand 3.77x10-5 failure during operation - transfers 3.63x10-6 2.07x10-6 i
closed (per hour)
RHR_Ritt_Exchangern failure during operation - excessive 1.15x10-6 1.02x10-6 leakage or fouling (per hour)
RHR_Hoom_Coolera failure to operate on demand 1.45x10-3 1.31x10-3 failure during operation due to 2.69x10-6 1.63x10-6 excessive leaking or fouling (per hour)
CS Roos _Conlers failure to operate on dematid 1.21x10-3 4.41x10-3 failure during operation due to 2.69x10-6 2.69x10-6 excessive leaking or fouling i
(per hour)
Drytell_ Air._Compr.eanora failure during operation (per hour) 7.40x10-5 8.89x10-5 4
ElanLContral_ Air _Compreanorm i
failure during operation (per hour) 7.34x10-5 8.96x10-5 l
Control _ Air _ Dryers l
failure during operation (por hour) 3.55x10-5 3.34x10-5 l
Air _ Relief Valves 6.51x10-6 failure during plant operationt lifting prematurely (per hour) i f
i
- In the January 1986 version, data for stotor operated valve failure to operate i
on demand (mean valuet 3.28x10-3) was used.
TABLE 5-1 Pcg2 3 of 4 IMPACT OF CHANCES ON COMPONENT FAILURE DATA AND INITIATING EVENT FREQUENCIES II.
INIIIAIING_EVENI JAIA All values shown are the mean of a distribution.
Annuni Frequency (per calendar _ rear _1 Internal Initiations Jant.ary 1986 September 1987 Erent_ Category (cyent_ Hod Version _
versj on_
Feedwater Rampup (1) 2.26x10-1 1.23x10-1 Moderator Temperature Decrease (2) 3.32x10-1 1.89x10-1 MSIV Closure (3) 9.54x10-1 7.50x10-1 Loss of Condenser Vacuum (4) 3.47x10-1 2.36x10-1 Loss of Of f site Power (5) 6.10x10-2 3.86x10-2 Pressure Regulator Failure-Closed (6A) 3.91x10-1 2.82x10-1 Pressure Regulator Failure-Open (6B) 6.35x10-2 4.16x10-2 Other Turbine Trip (7) 2.93 1.68 Loss of Feedwater* (8) 1.08 Loss of Feedwater-Not Imediately 4.93x10-1 Restorable * (8A)
Loss of Feadwater-Imediately 1.20x10-1 Restorable * (8B)
Partial Loss of Feedwater* (80) 1.48x10-1 Loss of Control Air (9) 1.56x101 9.66x10-2 Other Scram (10) 4.16 3.40 Main Steam Line Break-Outside 6.16x10-5 3.32x10-5 Containment (11A)
Main Steam Line Break-Inside 7.15x10-5 3.86x10-5 Containment (11B)
Feedwater Line Break-Outside 9.40x10-5 5.07x10-5 Containment (12/ )
Feedwater Line Break-Inside 4.36x10-5 2.35x10-5 Contain e.t (12B)
IIPCI Steamline Break (13) 3.30x10-5 1.78x10-5 RWCU Break-Return Line (14A) 7.18x10-5 3.87x10-5 RWCU Break-Suction Line (14B) 1.08x10-4 5.80x10-5 RCIC Lteamline Break (15) 4.29x10-5 2.32x10-5 Recirculation Discharge Line Break (16) 3.07x10-4 1.65x10-4 Recirculation Suction Line Break (17) 9.13x10-5 4,91x10-5 Core Spray Line Break-Inside 8.21x10-5 4.43x10-5 Containment (18)
T-15
TABLE 5-1 Pcg3 4 ef 4 IMPACT OF CHANGES ON COMPONENT FAILURE DATA AND INITIATING EVENT FREQUENCIES
- /anual Frequency (per cal.gndar_ysatl Internal Initiations January 1986 September 1987 EYanLCatsgory_fevent Nad Verslon Version _
Medium Steam Break I - InsHe 6.86x10-6 3.70x10-6 Containment (19A)
Medium Steam Break II - Inside 7.'46x10-6 4,03x10-6 Containment (198)
Medium Steam Break - Inside 4.53x10-5 2.44x10-5 Containment (190) l Small Steam Break - Inside 2.42x10-3 1.71x10-3 Containment (20A)
Small Water Break - Inside 4.04x10-2 2.53x10-2 Containment (208)
Inadvertent Opening of 1-3 MSRVs (21A) 1.13x10-1 6.58x10-2 Inadvertent Opening of 4 or more 1.96x10-3 1.41x10-3 MSRVs (21B) o
- Loss of Feedwater divided into three subcategories for the September 1987 version.
i I
l l
r 6
t i
T-16
o TABLE 5-2 Page 1 of 3 IMPACT OF CHANCES ON SYSTEM UNAVAILABILITIES January 1986(1) September 1987(1)
System / Top Event Version Version Residual Heat Removal System E1: Auto start of 2 RHR pumps / torus cooling 9.20x10-4 2.82x10-4 R2: Auto start pf 3 RHR pumps / torus cooling 1.83x10-2 1.34x10-2 and injection RS: Auto start of 2 RHR pumps (different loops)/ torus 3.98x10-2 3.32x10-2 cooling (High Drywell Pressure)
R7: Torus cooling 2.82x10-4 1.03x10-4 RB: Auto start of 2 RHR pumps / torus coolinr,(High 1.96x10-3 Note 2 Drywell Pressure)
RD: Auto start of 2 RHR pumps / torus cooling (multiunit) 1.96x10-3 1.03x10-3 SD: Shutdown cooling 1.93x10-2 1.99x10-2 SP: Containment spray 9.90x10-3 Note 3 Core Spray System C1: Auto start of I core spray loop (High Drywell 5.29x10-4 6.01x10-4 Pressure)
CS:
Auto start of I core spray loop 5.29x10-4 6.0x10-4' HPCI/RCIC*
H3: Auto start of HPCI or RCIC (High Drywell Pressure) 1.48x10-2 9.57x10-3 F.' : Auto start of HPCI 1.15x10-1 1.06x10-1 HP: Auto start of HPCI or ROIC 1.23x10-2 8.15x10-3 Emerr,ency Equips.ent Cooling Water System EE: Auto start of 2 pumps 2.42x10-3 6.41x10-4 Main S'~eLJCondensate/Fedwater CO: Cot 55ensate/ condensate booster pumps 1.45x10-3 1.18x10-3 F1: Manaal reestablishment of feedwater 1.49x10-3 3.45x10-4 FT: Feedwater Pump Trip 7.96x10-3 7.83x10-3 EH: Turbine Control System 1.67x10-2 1.79x10-2 T-17 9
I
o r
e i
TABLE 5-2 Page 2 of 3 IMPACT OF CHANCES ON SLN UNAVAILABILITIES January 1986(1) September 1987(1)
System /To9 Event Version Version Depressurization/ Relief Valves VI: Manual Actuation of one Relief Valve 6.32x10-4 1.88x10-4 Y2: Manual Depressurization 1.40x10-3 3.31x10-4 Ml: Retief 7alves Rescat 3.41x10-2 6.69x10-2 Reactor Protection System T2: Scram (L3 or High Drywell Pressure) 1.25x10-4 1.12x10-4 T3: Scram (L3 or MSIV position) 4.47x10-4 1.12x10-4 Recirculation Puse Trip RP: Recirculation Pump Trip 1.41x10-4 6.51x10-3 Raw C3oline. Water CW: Raw Cooling Water 2.96x10-4 1.12x10-4 Electric Power System EPS16: Unit I and 2 4KV shutdown boards available 9.92x10-1 9.97x10-1 EPS1:
4KV shutdown board A unavailable 5.95x10-4 7.83x10-4 EPS2:
4KV shutdown board B unavailable 5.95x10-4 7.85x10-4 EPS3:
4KV shutdown board C unavailable 5.91x10-4 6.46x10-4 EPS4:
4KV shutdown board D unavailable 5.91x10-4 9.79x10-4 EPS5:
4KV shutdown boards A B unavailable 6.06x10-6 1.52x10-5 EPS6:
4KV shutdown boards A C unavailable 7.04x10-7 2.97x10-7 EPS7:
4KV shutdown boards A D unavailable 7.04x10-7 6.31x10-7 EPS8:
4KV shutdown boards B C unavailable 7.04x10-7 5.72x10-7 EPS9:
4KV shutdown boards B D unavailable 7.04x10-7 6.31x10-7 EPS10: 4KV shutdown boards C D unavailable 1.31x10-4 2.80x10-6 EPS11: 4KV shutdown boards A B C unavailable 2.10x10-7 3.41x10-7 EPS12: 4KY shutdown boards A B D unavailable 2.10x10-7 3.44x10-7 EPS13: 4KV shutdown boards A C D unavailable 4.13x10-9 5.91x10-10 EPS14: 4KV shutdown boards B C D unavailable 4.13x10-9 5.91x10-10 EPSLS: 4KV shutdown boards A B C D unavailable 3.10x10-9 1.57x10-10 SA: Early Recovery of Offsite Power 4.50x10-2 4.5cx10-2 5B: Backfeed of Unit Board 3.00x10-1 3.00x10-1 T-18 1
g 9
TABLE 5-2 Page 3 of 3 IMPACT OF CHANGES ON SYSTEM UNAVAILABILITIES January 1986(1) September 1987(1)
System / Top Event Version Version Primary Conteinment Isolation CI: MSIV Isolation-L2 or low steamline Pressure 5.04x10-4 1.50x10-4 GH: MSIV Isolation-L2 7.93x10-4 1.55x10-4 t
Plant Air i
AI: Plant Air 6.11x10-3 1.0dx10-3 Standby Liquid Control SL: Standby Liquid Control 1.80x10-2 1.81x10-2 Conunon Actuation Sensors **
All sensors Available 9.74x10-1 9.79x10-1
{
Minim m sensors ave.ilable 2.53x10-2 2.06x10-2 l
Minima sensors unavatlable 8.12x10-4 4.89x10-4 l
- In the September 1987 version, credit is taken for the CRD system where appropriate.
For the earlier versione, CRD was modeled as a separate top event. This caused the apparent increase in the HPCI/RCIC unavailabilities for the September 1987 version.*
- CAS unavailabilities are for the alpha model.
Notes l
1.
System unavailabilities are for CAS state 1 and electric power state 16, offsite power available.
l 2.
Top event RB was replaced with top event R7.
i 3.
Top event SP was deleted between the January 1986 version and the September 1987 version.
T-19 1
i f
i
e TABLE 5-3 Page 1 of 3 CHANCES TO SYSTEM IMPORTANCE i
Januat'y 1986 Septe W e 1987 System / Top Event Versia_n Version Residual Heat Removal System 26.55%
17.53%
R1: Auto start of 2 RhR puseps/ torus cooling 10.28 R2: Auto start of 3 RHR pumps / torus cooling 0.51 and injection RS: Auto start of 2 RHR puses (different loops)/ torus 0.27 cooling (High Drywell Pressure)
R7: Torus cooling 3.88 14.86 RB: Auto start of 2 RHR pumps / torus cooling (High 3.28 Drywell Pressure)
RD: Auto start of 2 RHR pumps / torus cooling (multiunit) 3.45 1.71 SD: Shutdown cooling 5.00 0.18 SP: Containment spray 0.66 Core Spray 1.971 C1: Auto start of I core spray loop (High Drywell 0.91 Pressure)
CS: Auto start of I core spray loop 1.06 HPCI/RCIC*
23.08%
30.24%
H3: Auto start of HPCI or RCIC (High Drywell Pressure) 0.51 H4: Auto start of HPCI 6.47 1.29 HP: Auto start of HPCI or RCIC 16.61 28.44 Emergency Equipment Cooling Eter 0.36%
5.12%
EE: Auto start of 2 pumps 0.36 5.12 Main Steam / Condensate /Feedwater 3.73%
5.53%
CO: Condensate / Condensate booster pumps 1.21 2.89 F1: Manual reestablishment of feedwater 2.52 FT: Feedwater Pump Trip 2.13 CH: Turbine Control System 0.51 T-20 t
4
v P
'l f
TABLE 5-3 Page 2 of 3 CHANGES TO SYSTEM IMPORTANCE January 1986 September 1987 System / Top Svent Version Version Deprescurization/ Relief valves 16.78%
25.36%
VI: Manual Actuation of One Relief Valve V2: Manual Depressurization 15.60 25.36 MI: Relief Valves Rescat 1.18 Reactor Protection System 1.44%
0.46%
T2: Scram (L3 or High Drywell Pressure) 0.46 T3: Scram (L3 or MSIV position)**
1.44 Recirculation Puno Trip 0.44%
RP: Recirculation Pump Trip Function 0.44 Raw Coolinr Water 0.31%
CW: Raw Cooling Water 0.31 Electric Power 70.77%
16.29%-
EPS16 Unit 1 and 2 4KV Shutdown Boards available 54.39 S1.31 EPS1 4KV Shutdown Board A unavailable 7.40 7.06 EPS2 4KV Shutdown Board B unavailable 0.25 EPS3 4KY Shutdown Board C unevallable 0.28 4.81 EPS4 4KV Shutdown Board D unavailable EPSS 4KV Shutdown Boards A B unavailable 1.33 4.35 EPS6 4KV Shutdown Boards A C unavailable 1.28 3.07 EPS7 4KV Shutdown Boards A D unavailable 1.44 0.96 EPS8 4KV Shutdown Boards B C unavailable 1.36 3.47 EPS9 4KV Shutdown Boards B D unavailable 1.25 0.82 EPS10 4KV Shutdown Boards C D unavailable 0.22 EPS11 4KY Shutdown Boards A B C unavailable 0.31 0.19 EPS12 4KV Shutdown Boards A B J unavailable 0.48 EPS13 4KV Shutdown Boar 6s A D C unavailable 0.46 EPS14 4KV Shutdown Boards B C D unavailable 0.30 EPS15 4KV Shutdown Boards A B C D unavailable 0.27 SA: Early Recovery of Offsite Power 7.28 7.23,
SB: Backfeed of Unit Board 0.43 I
T-21
o.
l I
TABLE 5-3 Page 3 of 3 CMABGES TO SYSTEM IMPORTANCE January 1984 September 1987 System / Top Event Version Version Primary Containment Isolation 0.92%
1.65%
CI: MS1Y Isolation (L2 or Low Steamline Pressure) 0.92 1.32 CH: MSIV Isolation (L2) 0.33 Plant Air 2.06%
AI: Plant Air 2.06 Standby Liquid Control 0.12%
1.79%
SL: Standby Liquid Control 0. 5.'
1.79 Corunon Actuation Sensors 70.8M 75.29%
All Sensors Available 61.51 72.23 Minimum Sensors Available Minimum Sensors not Available 9.36 3.06 Percent CHF Represented by top 100 Sequences 70.871 76.29%
Absolute CMF from Internal Events (per calendar year) 3.9x10-3 4.7x10-4
- In the September 1987 version, credit is taken for the CRD System where appropriate.
- Second input signal added between the January 1986 version and the September 1987 version.
General hotes:
See section 5.3 for definition of importance.
Standby Cas Treatst System not included (SBGT only determined plant damage state).
Internal events at,y.
Importance values less than 0.01 are indicated by" T-22 8
i
o % e TABLE 5-4 P:ga 1 of 2 IMPORTANCE OF INTERNAL INITIATING EVENTS Internal Initiating January September Event Calagnry__(event..ngd 1986 Version 1987 Version
(%)
(%)
Feedwater Rampur (1) 1.30 1.12 Moderator Temperature Decrease (2)
MSIV Closure (3) 3.62 8.01 Loss of Condenser Vacuum (4) 1.51 3.12 Loss of O!folte Power (5) 11.10 12.36 Pressure Regulator Failure - Closed (6A) 1.95 3.90 Pressure Regulator Failure - Open (6B) 1.62 2.67 Other Turbine Trip (7)
Loss of Feedwater**(8) 4.62 Loss of Feedwater - Not Immediately 6.74 Restorable **(8A)
Loss of Teedwater - Immediately Restorable *(8B)
Partini Loss of Feedwater*(80)
Loss of Control Air (9) 7.26 2.47 Other Scram (10) 0.57 0.52 Main Steam Line Break - Outside Containment (11A)
Main Steam Line Break - Inside Containment (118)
Feedwater Line Break-Outside Containment (12A)
Feedwater Line Break-Inside Containment (125)
HPCI Steam Line Break (13)
RWCU Break-Return Line (14A)
RWCU Break-Suction Line (14B)
RCIC Steam Line Break (15)
Recirculation Discharge Line Break (16) 0.47 Recirculation Suction Line Break (17) 0.27 Core Spray Line Break-Inside Containment (18) l l
l l
t T-23 l
. M e TABLE 5-4 Pego 2 of 2 IMPORTANCE* OF INTERNAL INITIATING EVENTS Internal Initiating January September Event catemory-(event no.)
1986 Vers (On 1987 Version
(%)
(%)
Medium Steam Break I-Inside Containment (19A)
Medium Steam Break II-Inside Containment (195)
Medium Water Break-Inside Containment (190)
I Small Steam' Break-Inside Containment (20A) 0.66 0.77
~'
Small Water' Break-Inside containment-(208) 5.05 7.03 Inadvertent Opening'of 1-3 MSRVs (21A) 7.08 1.95 Inadvertent Opening of 4 or more MSRVs (215)
Loss of Of fsite Power Resulting in 1-3 Stuck 0.36 11.95 Open Relief Valves Transient Resulting in 1-3 Stuck Open 13.33 8.30 Relief Valves Transient Resulting in Loss of Feedwater 2.97 1.54 Transient Resulting in Feedwater Rampup 1.94 0.80 Transient Resulting in Small LOCA Transient with subsequent Scram Failure 5.83 2.77 Percent CMF Represented by Top 100 Sequences-70.87 76.29 Absolute CMF from Internal Events 3.9x10-3 4.7x10-4 (per calendar year)
- Loss of Feedwater divided into three subcategories for the September 1987 version NOTE: See section 5.3 for definition of Importance.
L e
e e
T-24
e, C/ o TABLE 6-1 PIANT DAMAGE STATE (PDS) FREQUENCIES Eregurs.< (Per Calendar _ Year 1 January September 1986 1986 EDS Ysrsion Version _
1A 6.91x10-4 9.28x10-5 1D
- 3. 5'ir,10-4 3.72x10-5 IJ 2.31x10-6 4.26x10-7 IL 4.08x10-5 8.27x10-6 2A 4.65x10'd 8.51x10-6 2D 1.60x10-5 2.46x10-6 2L 2.12x10-6 1,30xto-7 3A 7.3]x10-5 1.24x10-7 3D 1.98x10-5 3.36x10-8 5A 1.44x10-4 2.17x10-6 5D 4.05x10-5 6.38x10-7 6A 3.87x10-5 1.17x10-7 6D 1.05x10-5 3,oixto-8 6L 3.51x10-6 4.22x10-Il 7A 5.12x10-4 1.11x10-4 7D 1.39x10-4 3.90x10-5 7J 1.48x10-6 4,49x10-7 8A 1.38x10-5 4,71x10-6 8D 3.76x10-6 1.27x10-6 9A 3.44x10-4 3.10x10-5 9D 9.60x10-5 8.89x10-6 9J 1.00x10-6 1.27x10-7 10A 1.02x10-4 2.41x10-5 10D 2.77x10-5 6.53x10-6 10L 3.23x10-5 9.26x10-9 13/.
5.85x10-4
(.16x10-5 13D 2.81x10-4 2.37x10-5 13J
- 1. 9 7x10--6 2.79x10-7 15A 2.31x10-4 2.93x10-6 15D 6.42x10-5 8.16x10-7 TOTAL 3.9x10-3 4.7x10-4 NOTES:
1.
Plant Damage States are defined in Table 6-1 Part B.
1.
The values for the following Plant Damage States are < 10-6 for both the January 1986 and the September 1987 versions:
2B, 20, 2F, 2J, 3J, 3L, 4A, AB, 4C, 4D, 4F, 4J. 4L, SJ, SL, 6J. 7L, 8J. 8L, 9L, 10B, 100, 10F, 10J. 13L, 15J. 15L.
T-25
e af o
' TABLE'6-l'Pa'rt B DEFINITION OF PLANT DAMAGE STATES (PDS)
~
PDS Core Melt Time, Vassel Debris Bod
> 1 ft water
< 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> after Pressure Cooling in Drywell at (Numeric)
Event Initiation At Melt Availability Time of Melt
< 400 psai i
l' Yes Yes No No 2
Yes Yes No Yes 3
Yes Yes Yes No 4
Yes Yes Yes Yes 5
Yes No No No 6
Yes No No Yes 7
Yes No Yes No 8
Yes No Yes Yes l
9 No Yes No No
- 0 No Yes No Yes 11 No Yes 12 No Yes 13 No No No No 14
'No No No Yes 15 No No Yes No 4
16 No No Yes Yes i
PDS Primary, Containment Elevated Other Intact at time Release
- Filtering l
.(Alpha)
'of Melt Mechanisms I
A Yes No Yes SP l
B Yes SBGT l
C
'Yes None D
Yes No SP f
l E
Yes No F
Yes No None G
No H
No I
No J
No No SP K
No No l
L No No None I
SP = Suppression Pool
[
SBGT = Standby Gas Treatment Charcoal Filter, No SP i
'* Includes SGBT Roughing and HEPA Filters
- No Possible (No SBGT) or Forbidden by the Model 1
T-26
r TABLE 6-2 DQti1NANT SEQUENCES (Update of Table 2 of October 1,1987, NRC Letter)
EREQUENCY I
1.
Loss of Feedwater Transient, Failure of HPCI 2.2% E-5/ Year and RCIC, Failure of Manual ADS Blowdown, Electric Power State 16 CAS State 1 No' I
Recovery.
2.
Small LOCA, Failure of Torus Cooling, Electric 2.05 E-5/ Year Power State 16, CAS State 1, No Recovery.
3.
Main Steam Isolation Valve Closure, Failure of 2.04 E-5/ Year l
HPCI and RCIC, Failure of Manual ADS Blowdown, t
Electric Power State 16, CAS State 1, No Recovery.
4.
Transient with 1-3 Stuck Open Relief Valves, 1.58 E-5/ Year Failure of Torus cooling, Electric Power State 16, CAS Jtate 1. No Recovery.
5.
Loss of Offsite Power, Failure of HPCI and 1.25 E-5/ Year RCIC, Failure of LPCI Injection, Failure s
(
of Torus Cooling, Failure of Core Spray, Electric Power State 1, CAS State 1 Recovery.
l 6.
Pressure Regulator Fails Closed. Failure of 1.09 E-5/ Year HPCI and RCIC, Failure of Manual ADS Blowdown. Electric Power State 16, CAS State 1, No Recovery.
(
l 7.
Loss of Condenser Vacuum, Failure of HPCI 9.91 E-6/ Year and RCIC, Failure of Manual ADS Blowdown, Electric Power State 16. CAS State 1 No Recovery.
i r
8.
Loss of Offsite Power Resulting in 1-3 9.37 E-6/ Year i
Stuck Open Relief Valves Failure of EECW, l
Electric Power State 3. CAS State 1 L
Recovery 9.
Main Steam Isolation Valve Closure, Failure 8.67 E-6/ Year i
of HPCI and RCIC, Failure of the condensate System, Failure of Manual ADS Blowdown, Electric Power State 16. CAS State 1, No Recovery.
- 10. Transient with 1-3 Stuck Open Relief Valves, 8.09 E-6/ Year Failure of HPCI, Failure of RHR, Electric Power State 3, Recovery.
T-27
,